Digital Multimedia Information Retrieval Using Bootstrap Aggregative Learning Classifier

A digital multimedia information retrieval plays significant role in the field of image and video retrieval to retrieve the information relevant to a user query. In past decades, many research works have designed for digital multimedia information retrieval. However, classification performance of multimedia information retrieval was not efficient. In order to solve this limitation, Bootstrap Aggregative Learning Classifier (BALC) technique is proposed. The BALC technique is designed with improving the performance of digital multimedia information retrieval using machine learning classifier technique. The BALC technique initially takes video query as input and it applied Kernel-principal component analysis (Kernel-PCA) for extracting the visual features such as shape, color, texture in videos. After that, BALC technique used Bootstrap Aggregation with Support Vector Machine (BA-SVM) Classifier to classify the videos in a given dataset as relevant or irrelevant using video query with improved classification accuracy. At last, the BALC technique retrieves classified relevant videos' based on video query. This process assists BALC technique to improve precision and recall of video retrieval with minimum time. The BALC technique conducts experimental work on parameters such as classification accuracy, time complexity, precision and recall using three datasets with higher classification accuracy and minimum time complexity for multimedia information retrieval as compared to state-of-the-art works.